During the roundtable discussion, Dr. Xiao Feng mentioned that medical data has immense value, but if computation cannot be performed while protecting privacy, hospitals will be hesitant to share their data, hindering the practical application of this data in drug development and treatment planning. Privacy-preserving computing technologies such as fully homomorphic encryption and zero-knowledge proofs address the problem of "data usable but not visible."

However, he believes that AI + privacy computing alone is essentially still a Web2 business model: it requires negotiation, signing agreements, obtaining licenses, and then paying. The real breakthrough lies in adding blockchain, transforming data into callable, billable, permissionless, and trustless AI tokens. In this way, hospitals wouldn't need to know who the caller is; as long as the data is used, they would automatically generate revenue, significantly expanding the boundaries of business.